Google DeepMind
Updated May 2026
Gemini
Google's answer, built natively multimodal
Gemini is the merger of Google Brain and DeepMind (the lab that built AlphaGo) into one research effort. Unlike Claude or GPT, Gemini was trained from the start to handle text, images, audio, and video together rather than bolting vision onto a text model. It's also the only frontier family with the deep search, Gmail, Docs, and YouTube integration that only Google can ship.
Variants
9
Best at
massive context windows (up to 1M)
Max context
1M
tokens
Pricing
Free
$0
per Gemini app with 2.5 Flash
Pro
$19.99
per month with Gemini 3.1 Pro and 1000 AI credits
Ultra
$124.99
per 3 months with Deep Think and 25000 credits
Gemini 3.1 Pro API
$2 in / $12 out
per million tokens (<200k)
Gemini 2.5 Pro API
$1 in / $10 out
per million tokens
Variants
Sort the table by context window or cost to find the right variant. Click any version below for a battle card with ranks, pricing notes, and official links.
| Modalities | ||||
|---|---|---|---|---|
Gemini 3.1 Pro gemini-3-1-pro | 1M | varies / varies | 2026 | textvisionaudiovideocode |
Gemini 3 Pro gemini-3-pro | 1M | varies / varies | 2025 | textvisionaudiovideo |
Gemini 3 Deep Think gemini-3-deep-think | 1M | varies / varies | 2025 | textvision |
Gemini 2.5 Pro gemini-2-5-pro | 1M | $1.25 / $10 | 2025 | textvisionaudiovideo |
Gemini 2.5 Flash gemini-2-5-flash | 1M | $0.30 / $2.50 | 2025 | textvisionaudio |
Gemini 2.5 Flash-Lite gemini-2-5-flash-lite | 1M | varies / varies | 2025 | text |
Gemma 3 gemma-3 | 128K | varies / varies | 2025 | textvision |
Imagen 4 imagen-4 | — | varies / varies | 2025 | image |
Veo 3.1 veo-3-1 | — | varies / varies | 2026 | videoaudio |
Battle card
Context rank
#1
within Gemini
Capability rank
#1
modalities + reasoning
Weights
Closed
API/app access
Best fights to pick
- analyzing hour-long videos
- multi-hour meeting summarization
- cross-document research
- Google Workspace automation
Rankings are Tendril directory ranks, computed from the model data shown here. Public benchmark leaderboards change often, so official docs and current benchmark pages should be checked before buying or deploying.
Learn
Lessons about this model
Structured lessons that cover Gemini directly or put it in context alongside its rivals.
Gemini 2.5 Pro — how a 1M context actually helps
Gemini 2.5 Flash — free-tier use cases
Gemini Ultra — enterprise context windows
Gemini Deep Research — autonomous research pipeline
NotebookLM Audio Overviews — your papers, as a podcast
NotebookLM: Google's Source-Grounded Study Buddy
Claude vs. ChatGPT vs. Gemini — Side-by-Side
Check yourself
Quizzes
Short, mixed-difficulty quiz sets on Gemini and its model family.
Google's Gemini Family
5 questions
Gemini Pro, Ultra, Flash, Nano, and related Google AI models.
Start quiz →Hands-on
Try these prompts
Ready-made prompts that show Gemini at its best. Use them in your own AI workspace, then compare the output with what you learned in Tendril.
1M-token context with a long document
BuildersPaste a long transcript or essay and ask Gemini to find specific quotes — it shines when the haystack is huge.
I'll paste a 30-page meeting transcript below. Find every place where someone commits to a deliverable with a date, and give me the exact quote plus the deliverable owner. If dates are ambiguous ('next week'), flag them as ambiguous.
[Paste transcript here]Gemini Deep Research prompt
CreatorsGemini's Deep Research mode runs a multi-step research plan; give it a meaty question.
Research the current state of solid-state battery commercialization as of 2026. I want: (1) which companies have working production lines, (2) which are still lab-scale, (3) the realistic cost/kWh trajectory for the next 3 years, (4) the top 2 unresolved technical risks. Cite sources.
NotebookLM as a study buddy
ExplorersUpload your class notes and PDFs, then generate an audio overview — commute-friendly revision.
Using only the sources I uploaded, create a 10-minute audio overview that teaches the key concepts as if I'm a high-schooler hearing this for the first time. Avoid jargon unless you define it.
Multimodal: describe a chart
BuildersGemini's vision is strong on charts and diagrams — try handing it a screenshot of a business dashboard.
[Attach a chart image] Summarize this chart in three bullets: the headline finding, one nuance that casual readers would miss, and one question this chart DOESN'T answer but you'd ask next.
Print & keep
Printable reference
One-page summaries and flowcharts — great for desks, classrooms, or study sessions.
Go deeper
Official resources
Straight from the lab — docs, API references, and the chat surfaces you can try today.
Strengths
- massive context windows (up to 1M)
- native multimodality including video
- tight Google Workspace integration
Limits
- refuses more often than Claude or GPT
- product names keep changing
- Gemma is capable but not frontier